For dominance function,

x: A species abundance vector, or matrix with the absolute count data (no relative abundances).

This seems intended for 16S data analysis. How can a user calculate Simpson index for MetaPhlAn proportions, for instance?

1. Note that Simpson's dominance index (\lambda) is a different index than inverse Simpson (1/\lambda) or reciprocal Simpson (1-\lambda). The latter two are used as diversity indices but sometimes these 3 are confused.

2. Any of these indices is applicable to ecological communities regardless of the measurement technique. This can be taxonomic profiles derived from 16S or metagenomic experiments, or couting animals in a forest. Doesn't matter, really, as long as you are dealing with relative abundances of some classes.

3. Simpson's index is defined as $sum(p^2)$, where p are relative abundances. Metaphlan gives you relative abundances, so you can use the function microbiome::dominance with taxonomic abundance matrix that has been derived from metaphlan.

4. We do still minimal maintenance on the phyloseq-based microbiome R/Bioconductor package but since the last 3-4 years we ceased development of that, and moved on to developing R/Bioconductor microbiome tools centered around the (Tree)SummarizedExperiment data container; the mia package has functions for dominance and diversity as well, and importers for metaphlan data.